What do Models Learn From Training on More Than Text? Measuring Visual Commonsense Knowledge.

Lovisa Hagstrm, Richard Johansson

Browse the full ACL paper archive.